Aphasia secondary to stroke is a significant medical problem with potentially devastating psychosocial repercussions. Knowledge of the nature and time course of brain mechanisms that contribute to recovery of language function after stroke may have significant import for the development and evaluation of rehabilitative strategies for aphasia. In this regard, magnetoencephalography (MEG), the newest of the non-invasive functional imaging methodologies, provides the opportunity to characterize changes in language-specific neuronal activation during the course of recovery from aphasia, and correlate these changes with behavioral deticit. Combining excellent spatial and temporal resolution, MEG provides an important compliment to modalities that image changes in metabolism and hemodynamic response. Accordingly, the broad long-term objectives of this proposal are to: (t) Investigate the mechanisms of reorganization of language-specific cortex after aphasia secondary to first-time, unilateral, ischemic stroke in the left hemisphere using MEG; (2) Assess the efficiency of the reorganized mechanisms to support recovery of language; and (3) Establish the extent to which intensive constraint induced therapy for aphasia modulates the mechanisms of reorganization and their relationship to recovery. We propose a longitudinal study in which 100 patients with deficits in speech fluency and/or auditory comprehension, recruited over a 5-year period, will be followed for approximately 1 year each. MEG functional images will be obtained during the performance of receptive and expressive language tasks at 2 to 3 months and 10 to 12 months after stroke using protocols that have been shown to provide valid spatiotemporat maps of language-specific neuronal activity. Structural MRI images and comprehensive measures of speech fluency and auditory comprehension will be obtained concurrently with MEG functional images. These data will be used to address the following Specific Aims: I. To determine the effects of lesion volume and location on the spatiotemporal parameters of language-specific MEG activation. II. To determine the relationship between the spatiotemporal parameters of language-specific MEG activation and language function at 2 to 3 and 10 to 12 months after stroke; Ill. To determine the relationship between changes in the spatiotemporal parameters of language-specific MEG activation and changes in language function between 2 to 3 and 10 to 12 months after stroke. In addition, a subgroup of 40 patients will be administered Constraint Induced Language Therapy (ClLT), a novel therapy for aphasia, after the 10 to 12 month MEG imaging session. A Wait-Listed Control Design is used in which therapy is administered on a staggered schedule for half of the patients, allowing for a control group without denying therapy to any participants. All patients undergo MEG imaging and language testing both prior to and after the 3-week therapy session, This data will be used to address Specific Aim IV: To determine the changes in the spatiotemporal parameters of language-specific MEG activation after ClLT and test the hypothesis that these changes mediate the improvement in language function after therapy.